happy-app-audit

Warn

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes several system utilities including strings, codesign, and plutil through a whitelisted wrapper to extract app metadata and content.
  • [DATA_EXFILTRATION]: Accesses sensitive local directories such as ~/Library/Application Support and ~/Library/Logs to inventory files and analyze SQLite schemas related to app reporting.
  • [PROMPT_INJECTION]: Presents a surface for indirect prompt injection by processing untrusted strings from third-party binaries and presenting them to the agent for report generation.
  • [EXTERNAL_DOWNLOADS]: Uses npx to download and execute the bun runtime if it is missing from the host environment.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 20, 2026, 12:44 PM